paper-with-me

홈 › Papers

Orthogonalising gradients to speed up neural network optimisation

2022-02-14 · Mark Tuddenham, Adam Prügel-Bennett, Jonathan Hare

The optimisation of neural networks can be sped up by orthogonalising the gradients before the optimisation step, ensuring the diversification of the learned representations. We orthogonalise the gradients of the layer's components/filters with respect to each other to separate out the intermediate representations. Our method of orthogonalisation allows the weights to be used more flexibly, in contrast to restricting the weights to an orthogonalised sub-space. We tested this method on ImageNet and CIFAR-10 resulting in a large decrease in learning time, and also obtain a speed-up on the semi-supervised learning BarlowTwins. We obtain similar accuracy to SGD without fine-tuning and better accuracy for na\"ively chosen hyper-parameters.

📄 PDF Abstract BibTeX arXiv:2202.07052

Code (1)

MarkTuddenham/Orthogonal-Optimisers 공식 구현 pytorch

Methods 이 논문이 사용한 방법론

SGD Stochastic Gradient Descent is an iterative optimization technique that uses minibatches of data to form an expectation of the gradient, rather than the full gradient using…

Similar Papers 제목 키워드 기반

Orthogonalising gradients to speedup neural network optimisation

2021-09-29 · Mark Tuddenham, Adam Prugel-Bennett, Jonathon Hare

The optimisation of neural networks can be sped up by orthogonalising the gradients before the optimisation step, ensuring the diversification of the learned representations. We hypothesize that components in the same la…

Joint Sampling and Optimisation for Inverse Rendering

2023-09-27 · Martin Balint, Karol Myszkowski, Hans-Peter Seidel, Gurprit Singh

When dealing with difficult inverse problems such as inverse rendering, using Monte Carlo estimated gradients to optimise parameters can slow down convergence due to variance. Averaging many gradient samples in each iter…

Inverse Rendering

Explaining Process Control Optimisation Recommendations via GradientSHAP and Implicit Differentiation

2026-07-16 · Paul Darm, Cem Alpturk, Kenneth Ulrich, William Duncan 외 arxiv

Automated optimisation is increasingly adopted in industrial processes, yet a trust gap persists between engineers who design these algorithms and operators who must act on their recommendations. Explainable AI methods l…

signSGD: Compressed Optimisation for Non-Convex Problems

2018-02-13 · ICML 2018 7 · Jeremy Bernstein, Yu-Xiang Wang, Kamyar Azizzadenesheli, Anima Anandkumar

Training large neural networks requires distributing learning across multiple workers, where the cost of communicating gradients can be a significant bottleneck. signSGD alleviates this problem by transmitting just the s…

FastReg: Fast Non-Rigid Registration via Accelerated Optimisation on the Manifold of Diffeomorphisms

2019-03-05 · Daniel Grzech, Loïc le Folgoc, Mattias P. Heinrich, Bishesh Khanal 외

We present an implementation of a new approach to diffeomorphic non-rigid registration of medical images. The method is based on optical flow and warps images via gradient flow with the standard $L^2$ inner product. To c…

Optical Flow Estimation